I just want to get you input on the following idea. Until i have a airspeed pressure sensor i have decided to not attempt to control speed using SOG from the GPS. Instead my idea is do define a fixed "cruising power" that gives my plane a safe cruising speed at level flight.

To control altitude i use the GPS altitude and set pitch angle to get to the correct altitude. My idea is simply to add or subtract power to the engine in relation to pitch angle to keep the speed within safe limits.

power= cruise_power + pitch_angle * power_constant

It's a simple approach, but i think it could work until i can measure correct airspeed. What do you think?


Views: 410

Reply to This

Replies to This Discussion

I think wind will be a problem.
Why would wind be a problem? As long as i am running the plane at a fixed power setting it will move through the air at a given airspeed regardless of wind direction. Ground speed will off course vary depending on wind strength and direction. The throttle adjustments is just made to keep the airspeed during climbs and descents.

The method you describe is exactly what we use with ArduPilot. It works...okay. Maybe our planes aren't trimmed well enough, but they tend to fly superfast and lose a bit of altitude going downwind and practically come to a dead stop going upwind. That's why we're adding elevator altitude control with ArduPilot 2.0
Chris, Perhaps I'm missing something in your last post...

A lot of the foam electrics that I have seen don't do well in much wind, they don't have the power to fly into stronger winds. At our field the foamy fliers tend to pack it up when the wind picks up over say 8 to 10 mph. How will adding elevator control to the ArduPilot 2.0 give you more wind penetration? Why not a bigger motor?

Are you thinking that using a PID with the elevator will trim the wing to an optimum incidence angle for better wind penetration? Won't ArduPilot 2.0 still need throttle control to manage and maintain altitude?

Hi all guys,
I would like to take opportunity in this topic to ask you if somebody knows the topic in which another user time ago posted a useful link on mechanics equation regarding flight.
In the link also a lot of information on how to implement your flight model .
Any search trough the forum could help me...
Anybody knows the link I´m talking about?

Yes, optimum incidence angle. We may add throttle control on top of that.

Think of it this way: when you're flying manually with a foamy and you turn into the wind, it has a tendency to pitch up. If you kill the power, as a throttle-only autopilot would, it will try to wheel around and go downwind. You can fight that with the rudder, but we're just updating the rudder at 1Hz and our GPS directional vector is unreliable if the speed over ground is near zero, so it gets very unstable in that situation. Instead, you should keep the power going and use the elevator to change the pitch so it can make headway.

Ok, optimum incidence to penetrate into the wind, got it.

As you said, in the upwind state 1hz GPS update is too slow to sense a change in directional vector. Can you use data from the FMA XY sensor to measure the planes' angle of incidence? Are you are thinking of using some experimentally found plug-in value to set the elevator?

When I flew my RC glider on windy days we added weight at or near the CG for better handling. Maybe you can add a heavier camera or more gear!
"Can you use data from the FMA XY sensor to measure the planes' angle of incidence? Are you are thinking of using some experimentally found plug-in value to set the elevator?"

Yes, that's what ArduPilot 2.0, which has the FMA sensor integrated, does.
Pressure sensors vs. ir sensors.

in particular the rudder problem. As Chris points out, being turned back by the wind is a fairly significant challenge for UAV. IR can manage attitude and roll, but ir cannot help with yaw. a magnetic compass could, but that is twice the price of an ArduPilot for a single sensor. Gyros could also help, but again, increase the cost of the autopilot by a few multiples.

Enter pressure sensors. as cheap as $1 per device (plus amp). Pressure can regulate attitude and speed quiet nicely, but could it not also help regulate yaw or direction? It turns out that sailboats use a kind of autopilot which keeps them "faced into the wind" at just such an angle.

So - a theory here - with two pressure sensors fixed at gentle angles left and right of center, it should be possible to sense the wind speed and angle. If one could sense wind angle, one should be able to hold wind angle. I like the idea of an accelerometer/ dual pressure/gps. if the pressure sensors have a slightly downward bias (effective dihedral) they might provide a kind of roll-stabilization.

Anyone else think that multi-port pressure-based stabilization is interesting as cheap and good?

By "power" do you mean thrust, or propeller rpm relative to an indexed servo position? Or are you talking about electrical power (I don't know anything about the electric engines) related to rpm, and calculated thrust from propeller size/pitch? What you would like to do for maintaining altitude, of course, is to keep lift constant, and that is obtained by a combination of thrust and angle of attack (pitch angle) in coordinated flight (what I think you're suggesting). It sounds like what you are going for is: if the state engine senses a departure from the target GPS altitude, how does lift need to be modified to get there? Some other questions you need to ask yourself are:

- How fast do I need the aircraft to converge to the target altitude? If you merely use full control, you'll induce a pitch oscillation - so you'll want some sort of dampening function.
- What amount of error is acceptable? Else you will continue to chase your target altitude.

I think you are on the right track to ignore SOG.


I can't help but to make a few (not necessarily constructive) comments here. First, it's funny to hear that a 1Hz GPS is "too slow" when the project I worked on 9 years ago was using a .1Hz GPS. You'd think one order of magnitude would matter. In a few more years we'll hear that 10Hz is too slow. Second, it's just bad luck that as we scale the aircraft down, those pesky air molecules don't change size accordingly, and drag becomes a bigger problem. Third, related to drag, it's the glider RC pilots (and full scale glider pilots) that seem to understand the airspeed expense of uncoordinated flight the best.

presumably, years ago, you had real gyros, and a pressure altimeter. Given those of course, a 10 sec GPS is more than sufficient. What you have here is a bid to rely increasingly on the GPS, while the altimeter, and gyros have been diminished or eliminated. A Gyro autopilot can be had for $300, while ArduPilot is like $24. I would bet your autopilot of 9 years ago probably had an insurance value of several K.

As for scaling, small birds react much quicker, so the reaction time of their stability system must be that much faster.

As for coordinated flight - I see your point, you're suggesting that a simple rudder-based uav will experience more drag than a properly executed turn. You may note that several pilots here are using flying wing, and those can only make coordinated turns, so I'd have to think the Ardupilot can do coordinated turns.

Reply to Discussion


© 2017   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service